Multiphoton Microscopy has become the method of choice for intravital imaging at submicron resolution. It works by both temporally and spatially compressing very high numbers of near infrared photons into the focus of a microscope objective. Millimolar photon densities permit the simultaneous absorbtion of two photons by the fluorescent dye, yielding the same excited state one would get with a single bluer photon. This occurs only in a privileged (high photon concentration) zone about a micron tall and 250 nm wide, ellipsoidal in shape, known as the PSF (point spread function). Thus the tiny spot IS the image;one must simply raster it about to get a picture. Importantly, ALL light leaving the dye is useful. In confocal and/or camera based microscopes, only the light coherently imaged onto a detector is of value. In MPM, light can be collected in a "non-imaging" device and the computer reconstructs the picture from raster intensity. Unfortunately, conventional objectives recover only a small portion of the emitted light. The theoretical maximum in clear media is about a third for oil immersion, about a fifth for water objectives and only a tenth in air. In turbid media like tissue, these inefficiencies can double or triple in severity. We have designed and patented TED ("Total Emission Detection") devices to overcome these signal limits. First, in TEDI, we designed a device for cells and tissue blocks that increases typical signal levels an order of magnitude. In published accounts, we show the gain could be used to scan 9x faster or reduce laser power 3x to avoid photodamage. Most recently, in TEDII, we designed a device class that can approach living animals. In our published accounts, we show that although half the light is necessarily lost in the animal, we efficiently recover the rest, seeing e.g. 2.5x more light from the exposed rat brain. Again, this means we can either scan faster or reduce laser power a third. We are currently collaborating with a company to refine and manufacture TEDII devices , in order to quickly disseminate the technology to others. In the last year, we have also designed and tested adaptive optics (e.g. deformable mirrors) to compensate for the inhomogeneity of tissue. Just as astronomers must look though an inhomogeneous, moving atmosphere, we must generate the focal spot in translucent tissue. Both problems cause blur and twinkling. The solution is to use a deformable optic to compensate for the known distortion. In astronomy, a known "guidestar" point can provide that;in tissue, we must either build guidestars that stand out from the tissue reflection or use the computer to make succesive guesses to clean up the image. For the former, we are synthesizing our own multiple-layer nanoparticles to provide a separable clean "guidestar" signal inside tissue. We also begun tests of the contrast potential of these particles in an OCT (Optical Coherence Tomography) setting. More effort this year was devoted to mirror nonidealities. We have also evaluated the possibility of two-photon phosphorescence lifetime imaging for intracellular O2 detection, building a 2p and single photon microphosphorimeter. In addition to device development, we employ the multiphoton microscope to do FCS- Fluorescence Correlation Spectroscopy - of labeled molecules inside living cells. With FCS, we can count a few hundred transcription factors in the cell nucleus and determine their mobility (i.e. are they free or chromatin-bound?) and learn the role of cofactors. For example, we are studying the oncogene product C-myc and learning how its chromatin affinity is potentiated by its partner , MAX. Knockdowns and siRNA reveal more mobile C-myc. FCS can also be used to study protein-protein interactions throughout the cell. We previously published accounts of the promiscuous avidity of the HIV-1 protein Nef for cell-surface proteins like CD4 (reducing superinfection) and HLA-I (compromising immune surveillance), showing the affinity is maintained in several internal organelles. We have also identified (with collaborators) a bleaching artifact in FCS of very large (ca. 100MD) assemblies and published methods of correction/cross-validation. We continue evaluating this artifact in aggregating (e.g., plaque-forming) proteins.